What SaaS SEO is, and why it is a different job
SaaS SEO is search optimisation built around a software product rather than around a content site. Instead of chasing traffic, it targets the queries a buyer runs while evaluating tools — category terms, competitor comparisons, “alternatives to X”, integrations and jobs-to-be-done — and connects them to the trial or the demo.
That sounds like a difference of emphasis. It is not. It changes the build order, the page formats, the technical work and the report, because four things are true of software search that are not true of a blog.
- The buyer is comparing, not learning. They arrive mid-evaluation, typing a competitor’s name. A blog archive has nothing to say to them.
- The page is not the product. On a content site the page is what was bought. Here it is a route into a trial, so offer placement matters as much as position.
- The architecture breaks crawlers in specific ways. App subdomains, a separate docs platform, JavaScript-rendered marketing pages and parameterised URLs are all normal in SaaS and all hostile to indexing.
- The unit of success is revenue. Sessions are a diagnostic. The number that decides whether the programme continues is organic-sourced trials, trial-to-paid rate and MRR.
| Content site SEO | SaaS SEO | |
|---|---|---|
| What the page is for | The page is the product — a read is the outcome | The page is a route to a trial — a signup is the outcome |
| Who is searching | Someone learning a subject | Someone mid-way through choosing a tool |
| Build order | Topical depth first, commercial pages later | Commercial layer first, topical depth after it converts |
| Hardest technical problem | Speed, duplication, thin pages | App subdomains, docs platforms, JS rendering, URL parameters |
| What the report measures | Sessions, rankings, impressions | Organic-sourced trials, trial-to-paid, MRR per cluster |
Working with software companies worldwide
Maxinium is based in Colombo and this service is not scoped to Sri Lanka. Software is bought in the same language, from the same shortlists, in every market we work in — which is why the rest of this site is deliberately local and this page deliberately is not.
Two practical things follow. The first is time zones, and they are not a footnote when the person doing the work is also the person on your calls. The table below is the honest version: Sri Lanka is GMT+5:30, we hold a 09:00–18:00 day, and we extend into the evening for the Americas rather than pretending the overlap already exists.
| Where you are | Your hours that overlap ours | Shared working time |
|---|---|---|
| India (GMT+5:30) | 09:00 – 18:00 | Identical working day |
| United Arab Emirates (GMT+4) | 09:00 – 16:30 | 7.5 hours |
| Central Europe — Germany, Netherlands (GMT+2) | 09:00 – 14:30 | 5.5 hours |
| Singapore and Hong Kong (GMT+8) | 11:30 – 17:00 | 5.5 hours |
| United Kingdom (GMT+1) | 09:00 – 13:30 | 4.5 hours |
| Australia, east coast (GMT+10) | 13:30 – 17:00 | 3.5 hours |
| US and Canada, Eastern (GMT−4) | 08:30 – 11:30 | 3 hours, from our evening |
| US and Canada, Pacific (GMT−7) | 08:30 – 10:00 | 1.5 hours, from our late evening |
Offsets are northern-summer values. The UK, EU and North American rows shift by an hour when those markets leave summer time — Sri Lanka does not observe daylight saving.
The second thing: who actually does the work
The reason an offshore agency is usually a false economy is layering. You buy a strategist, you get an account manager, and the work is executed by whoever is free that week. Nothing about being in Colombo fixes that — it only makes it cheaper to do badly.
So the arrangement here is the opposite one. The person who audits your site is the person who ships the strategy. Alston Antony runs the engagement directly, there is no account-manager layer to route questions through, and the roster is capped at the number of programmes one person can actually run. That is a real constraint and it is stated rather than hidden: if the calendar is full, you will be told so instead of being handed to someone else.
The day job is part of the same argument. Alston is Senior Digital Marketing Manager at Brainstorm Force, whose WordPress and AI software powers more than 7 million websites — which means the SaaS advice on this page comes from running search inside a software business at scale, not from having read about it.
The six stages of a SaaS SEO programme
Every engagement runs through the same six stages in the same order. The order is the method: the commercial layer is built before content compounds against it, and revenue measurement is wired up before anything is published, so month six is not an argument about whether it worked.
| Stage | The question it answers | What lands on your side |
|---|---|---|
| 01 · Demand | Is there enough buyer-generated demand to justify the spend at all? | A ranked opportunity map: which clusters convert, which are citation-only, which to ignore, and the realistic trial ceiling for each |
| 02 · Architecture | Which pages does a buyer actually read before choosing a tool? | A page map with the commercial layer in build order — category, vs., alternatives, integrations, use cases — each URL tied to the one query it exists to answer |
| 03 · Content | Is this the most complete, most extractable answer available? | Briefs and published pages that lead with the answer, carry real HTML tables rather than screenshots, and contain operational detail a competitor cannot restate |
| 04 · Authority | Would an editor — or a model — cite this page as a source? | Editorial links, review-site and directory placement, named authors on trust-critical pages, and a stated methodology behind any scoring |
| 05 · AI visibility | Is the product on the shortlist before a single blue link is clicked? | Citation share tracked per surface and per query, reported beside organic clicks rather than instead of them |
| 06 · Revenue | Which clusters produced trials and MRR — and what gets cut? | A monthly readout tied to signups and revenue per cluster, with an explicit list of what is being stopped |
The stages overlap in practice — stage 04 does not wait for stage 03 to finish — but nothing skips ahead of stage 01, because a programme that was never sized is a programme nobody can end.
Technical SEO for SaaS architectures
SaaS sites fail in ways content sites do not, and almost none of it shows up in a generic audit tool with every warning switched on. The marketing site, the app, the docs and the changelog are frequently four different systems on three different hostnames, each with its own idea of canonicalisation.
The work below is the part of an engagement most likely to produce movement inside the first sixty days, because it acts on demand that already exists rather than creating new demand.
- App and marketing hostnames competing for the same queries, or the app leaking indexable pages
- Documentation on a third-party platform that ranks instead of — and sometimes against — the pages that convert
- JavaScript-rendered marketing pages whose copy never reaches an AI crawler, which reads raw HTML and does not execute scripts
- Parameterised and faceted URLs multiplying one page into thousands of crawl paths
- Programmatic template pages that are thin at scale, and the decision about which of them should exist at all
- Trial, pricing and signup flows measured properly in GA4 so the SEO report can be joined to the CRM
AI search visibility for software buyers
Software is the category where answer engines arrived first. A buyer asks an assistant for the three best tools in a category and gets a shortlist before visiting any website — and if the product is not in that answer, the click never happens and no ranking report will explain why.
This is the part of the service with the most measurement behind it, because it is where the owned portfolio has been pushed hardest. ZPlatform.ai, an AI tools directory Alston runs, recorded 241,500 Bing Copilot citations in six months against roughly 2,500 organic Bing clicks — about 97 citations per click — across 1,640 grounding queries in five languages, with 52% of them on commercial-intent searches. One page alone accounts for 108,546 of those citations. The raw exports are on the case study page.
The method is the same one in stage 05 above: make the entity unambiguous, put the load-bearing claims in short blocks near the top of the page, get placed in the directories and review sources that models actually cite in your category, and track share of voice per query. Surfaces that cannot be measured — ChatGPT and Gemini publish no citation export — are reported as not audited rather than estimated, here and on every case study.
Where the SaaS experience comes from
Plainly: there is no published client SaaS case study on this site yet. Saying so is cheaper than the alternative, which is implying one through careful phrasing and being found out on the first reference call.
What there is instead is a portfolio Alston owns and operates, five properties of which are software or AI products competing in international English-language search — the same search we would be doing for you, against the same competitors, with the exports published in full.
| Property | What it is | What it evidences |
|---|---|---|
| ZPlatform.ai | AI tools directory, every tool hand-tested before it is ranked | 241,500 Copilot citations in six months against ~2,500 organic clicks — full case study |
| SaaSPirate | SaaS and AI deals platform, running since 2019 | 2,513+ verified deals listed, a 100K+ community and 30K+ subscribers built around software buying intent |
| EveryAny.one | Software discovery across 13 categories | The comparison and alternatives architecture stage 02 describes, built and maintained rather than theorised |
| AIToolsMarketer | Independent reviews of AI marketing tools | Scoring against a published methodology — the stage 04 test of reading as a source rather than an aggregator |
| SaaSAffiliate | 62 SaaS affiliate programmes compared on payout terms | Programmatic comparison pages kept deliberately small and maintained, instead of thin at scale |
These are owned properties, not client work, and are labelled as such everywhere they appear on this site. The case studies carry the raw Search Console and Bing Webmaster Tools exports, the dates and the methodology behind every figure.
How engagements work, and what they cost
Four shapes, because a seed-stage product with one marketer and a Series B with a content team do not need the same thing.
International engagements are quoted in USD against a scope, after a call. There is no rate card on this page because there is no honest one to publish: the scope of a SaaS programme varies by an order of magnitude with the size of the commercial layer that has to be built. The published LKR retainers are for Sri Lankan businesses and are not the pricing for this service.
- Blueprint sprint — fixed fee, fixed window. Demand map, page architecture, technical priorities and a 90-day plan. Yours to keep and hand to anyone, including another agency.
- Fractional SEO lead — a monthly retainer where Alston runs search as part of your team: strategy, briefs, reviews, and the reporting joined to your CRM.
- AI visibility sprint — entity, structure and citation-source work aimed specifically at getting the product into AI shortlists, with share-of-voice tracked per surface.
- Team enablement — training and systems for an in-house marketer who will own the programme, structured so the dependency ends rather than compounds.
Who this is for, and who it is not for
The honest filter matters more in SaaS than anywhere, because the payback period is long enough that a bad fit wastes two quarters before it is obvious.
- Good fit: a live product with paying customers, a category buyers already search by name, and someone internally who can ship page changes without a three-week queue.
- Good fit: a founder-led or lean team where one person owns marketing and needs a system rather than another vendor.
- Poor fit: pre-product-market-fit, where the category has no search demand yet and paid experiments will answer the question faster and cheaper.
- Poor fit: anyone who needs pipeline this quarter. Technical and on-page work commonly moves inside 30 to 60 days; compounding content growth is a three-to-six-month story and it would be dishonest to sell it as anything else.
- Poor fit: a programme where the deliverable is a published word count. We will decline it rather than take it.